Chip makers are hoping a new breed of microcontrollers (MCUs) will begin laying the foundation for a solution to one of the auto industry's most vexing design problems -- electronic complexity.
Using more cores, bandwidth, Flash memory, and onboard intelligence, the new MCUs could begin to chip away at the complexity issue, which is growing worse, even as suppliers struggle to make in-roads. "It's out of control," Brad Loane, auto body microcontroller product manager for Freescale Semiconductor, told Design News. "There's so much more lighting, so many new user interfaces, so many more sensors and actuators. It's getting so complex that it's driving up the weight and the cost."
Indeed, automotive electronics are now so pervasive that they account for at least 30 percent of the overall cost of a vehicle -- and some experts suggest that's a conservative figure. For hybrids, it could be as high as 45 percent. And, in a few years, hybrids could reach as high as 80 percent.
I disagree with the idea of being too centralized. Certainly there could be some consolidation of processing, but a lot of automotive systems lend themselves to distributed control.
For example, if you have electric assist power steering, would you really want all of the control lines to the H-bridge and the encoder lines from it running up to a central computer? It makes a lot more sense to have a local MCU connected via a bus to a main CPU unit.
It may sound alarming to say that there are 100 MCUs in a car, but I could see that figure easily being touted as a positive thing. It reduces wiring and the associated weight, cost and EMI potential.
The other thread - reducing the number of modules talking to those 100 MCUs, makes more sense. But still, there are a number of systems that need the highest priority and you can't really have multiple first priorities in a system controlling multiple sub systems. In a multi-tasking system, is breaking more important than steering?
I really don't want my car control computer to reboot as I'm driving down the highway. What is being done to ensure that none of the low priority functions can lockup and bring down the overall controller? It happens all the time on my personal computer.
Great disclosure! Really enrich my knowledge. Would you read as follow #Rajawebhost http://dipotajukutama.blogspot.com/2012/12/mau-bikin-website-hosting-murah-abizz-ke-rajawebhost-com-aja.html | #Iconia http://virgosphere.blogdetik.com/2012/12/29/iconia-pc-tablet-dengan-windows-8-acer-iconia-w510/ | #TemaBLOGdetik http://bloggerpriok.blogdetik.com/2013/03/20/gratis-dari-blogdetik-144-tema-blog-yang-siap-pakai/ | #BukuTarifHSCodeBTKI http://didsbukutarifkepabeanan2013.blogspot.com/.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.